We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Saccharide concentration prediction from proxy sea surface microlayer samples analyzed via infrared spectroscopy and quantitative machine learning
Summary
Researchers developed infrared spectroscopy combined with machine learning to analyze dissolved organic carbon in sea surface microlayer samples more efficiently. The sea surface microlayer is a critical zone where microplastics concentrate and interact with marine chemistry.
The physical and chemical properties of the sea surface microlayer (SSML) are dynamic and complex. With an enrichment of organics from dissolved organic carbon (DOC) and many mechanisms for their release into the atmosphere, high-throughput analysis of SSML samples is necessary. Collection of more detailed information about the SSML would enable greater understanding of the release of ice nucleating and cloud condensation particles and provide critical feedback for climate models. The work presented herein details an investigation to develop machine learning (ML) methodology utilizing infrared spectroscopy data to accurately estimate saccharide concentrations in complex solutions. We evaluated several machine learning approaches toward this goal. Support Vector Regression (SVR) models are shown to predict the accurate generalized saccharide concentrations best. Our work presents an application combining fast spectroscopic techniques with ML to analyze SSML chemistry more efficiently, without sacrificing accuracy or precision.
Sign in to start a discussion.
More Papers Like This
Saccharide concentration prediction from proxy sea surface microlayer samples analyzed via infrared spectroscopy and quantitative machine learning
This study developed infrared spectroscopy methods combined with machine learning to predict saccharide concentrations in proxy sea surface microlayer samples. Accurate quantification of dissolved organics in the ocean surface layer is critical for understanding their role in cloud nucleation, ice formation, and other climatological processes.
Saccharide concentration prediction from proxy-sea surface microlayer samples analyzed via ATR-FTIR spectroscopy and quantitative machine learning
Researchers investigated machine learning models for predicting saccharide concentrations in sea surface microlayer (SSML) samples analyzed by ATR-FTIR spectroscopy, evaluating support vector regression (SVR) and other approaches on proxy-SSML samples. SVR models achieved the best predictive accuracy for saccharide concentration, enabling high-throughput SSML analysis to better understand organic enrichment and climate-relevant aerosol particle release.
Microplastic particles in the Arctic marine environment: database of IR spectra and its analysis by machine learning methods
Researchers built a database of IR spectra from microplastic particles collected across Arctic marine environments and applied machine learning methods to enable faster and less labor-intensive chemical composition analysis, identifying polymer types from spectral signatures at broad regional scales.
Microplastic particles in the Arctic marine environment: database of IR spectra and its analysis by machine learning methods
Researchers compiled a database of infrared spectra from microplastic particles collected in the Arctic marine environment and applied machine learning methods to automate polymer identification, addressing the labor-intensive nature of manual spectral analysis. They developed and evaluated ML classification models using real environmental polymer spectra to improve the speed and scalability of microplastic chemical characterization in polar research.
Study on marine microplastics monitoring based on infrared spectroscopy technology
Researchers developed an infrared spectroscopy-based monitoring system for marine microplastics, applying support vector machine algorithms to hyperspectral images to identify plastic types and abundances in seawater. The study found microplastic abundances ranging from roughly 5 to 39 particles per litre across sampling sites, with fibers (53-68%) and debris (23-34%) as dominant shapes, demonstrating the method's feasibility for rapid environmental monitoring.